Online diagnosis and monitoring of roundness defects in CNC machining processes. (18th October 2010)

• Record Type: Journal Article
• Authors: Garcia-Plaza, Eustaquio; Nunez, Pedro J.; Martin, Angel R.; Beamud, Elena; , Carlos De la Cruz
• Is Part Of: International journal of mechatronics and manufacturing systems. Volume 3:Number 5/6(2010)
• Page Start: 357
• Page End: 367
• Publication Date: 2010-10-18
• Subjects: online monitoring -- regression -- cutting forces -- roundness errors -- online diagnosis -- CNC turning -- roundness defects -- CNC machining -- predictive modelling -- quality control -- fault diagnosis; Mechatronics -- Periodicals; 629.89
• Languages: English
• ISSNs: 1753-1039

Abstract

The main aim of this study was to design and develop an online diagnostic system for the real-time prediction of roundness error (ER) in automated CNC turning processes. Initially, during the first phase, cutting force signals (Fx, Fy, Fz) were registered during the machining process. Thereafter, the signals were characterised using a series of parameters, i.e., time domain and previously selected frequency. The parameters that were found to significantly influence roundness error, which constituted the basis of the predictive models using regression techniques, were analysed. This global monitoring system has been developed with Dasylab 9.0, which enables the register and characterisation of signals as well as the real-time prediction of the roundness error (ER). The system predicted the parameter ER with a mean relative error percentage of ∼10%, a degree of precision similar to that obtained by post-process quality control of roundness using form measuring.